We become scientists because we love to learn, so the prospect of teaching eager young minds--giving them their first glimpses into worlds we love--appeals to most of us on some level. It's also something we have to be able to do well--or at least competently--to get tenure at a good institution, whether the institution is focused on teaching or research.
Becoming a master teacher takes time, experience, confidence, and open-mindedness--especially experience, which you can't force. But, becoming a competent teacher is all that's usually needed to pass muster, and that isn't exactly rocket science -- unless, I suppose, you happen to be teaching rocket science.
The problem for most new faculty is not so much the difficulty of teaching. Teaching is hard, but most new faculty members are used to doing difficult things. The main problem is that there are so many other things that the typical faculty beginner is called upon to do.
In the interest of making this job--learning how to teach--a bit easier, I've laid out some guidelines to help you efficiently gain confidence and competence as a teacher. By the way, for teachers in training--and we're all in training--I also recommend Tara Kuther's article Teaching 101: Getting By .
What to Teach--The Forest
The first thing you need to figure out when planning a new course is what to teach. Teaching exactly what you learned when you took the same course is not always the best approach; some thought should be given to selecting content. The three best guides to deciding what to teach are your colleagues (either at your home institution or at comparable institutions), textbooks, and Web sites that disseminate curricular material.
If you look around online, you can find course materials for many different subjects. Don't swallow any of it whole--no matter what you teach you need to make the course your own--but a quick look at these sites (and other resources) can help you make wise choices as to what, and how much, to include. The Galileo project  at Harvard is designed to provide materials needed for developing new science courses. MIT provides what it calls "open courseware"  including downloadable lecture notes, syllabi, and problem sets for a large number of MIT courses, all free, no password required!!
How much is a question that you shouldn't overlook. By the time you get your first faculty position you likely will have come to believe that everything taught in the core courses in your field is essential; after all, you probably missed a few key points yourself and regretted it later. Yet it is a mistake--one of the most common mistakes in my experience--to try to do too much. If you try to teach too much material your students will end up learning less of it than if you try to cover less.
What to Teach--The Trees
Science educators debate whether courses--especially introductory courses--should be structured around content (e.g., the basic topics traditionally taught in a second-semester chemistry course) or context (e.g. AIDS, or biogeochemical cycles). Introductory textbooks are available ( Chemistry in Context, Biology in Context, Biology Today) that take the latter approach, often to good effect. Both approaches have their merits and supporters. Which you should choose will depend on what you are most comfortable with, what kind of institution you're teaching at--you might make one decision at MIT and a different one at, say, Oberlin--and what the colleagues who will be evaluating you for tenure think about the relative merits of the two approaches. Use your own judgment, but there's no need to alienate needlessly your likely executioners.
Whatever choice you make, recall that a key feature of science is that it provides tools for solving new problems. As a profession, science is not really about big ideas, if anything, it's the story of big ideas being discredited by careful experimentation, ruthless honesty, and extraordinary cleverness. Give your students skills, not just stories.
Consider your audience. For beginning teachers it is easy to forget how much knowledge we have--and how little our students have--because our knowledge of science has so profoundly shaped our worldviews. Though the best students are likely to make a valiant attempt, they can't hang with you if you don't meet them halfway--or in some cases 95% of the way. You'll need to think hard about--and plan well--how you'll pass along the essential (if sometimes implicit) knowledge of basic concepts that your students need but that you might not realize they lack unless you think about it. More advanced and more gifted students can absorb things, as it were, by osmosis, but for introductory students and less gifted students you must either teach those concepts explicitly or avoid them altogether. For the majority of students there is no middle ground.
How to Teach
Next, you need to think about how to teach. Most of us learn best when we do things; rote memorization is an inferior approach, though in many instances it's the only approach that's feasible; more on this later. The idea of learning by doing--along with a healthy desire to maintain a cheap scientific labor force--is the foundation of the research-based Ph.D., but the experiential approach works equally well for younger students, although you have to make adjustments for the lower motivation, less self-confidence, and greater dependence (on you, the teacher) of younger students.
"Experiential learning," as it is often called, results in better retention, but it is not practical or efficient to have every student in your introductory biology class do a research project on every topic that you'd like to cover. Somehow you need to find a way to balance the dual demands of transmitting the necessary information and engaging students' minds in ways that will lead to superior retention. The best solution is often a balance of lecturing and experience-based techniques.
There is no single way to teach. You must find a method that works for you, that builds on your strengths, that you feel credible using. Feeling credible is very important to your effectiveness as a teacher. As you teach, look for opportunities to experiment, to try new things. And remember that basic tenet of scientific practice: though we may prefer a positive outcome, failed experiments, too, yield knowledge.
What, in practical terms, are the teaching-related options? There's lecturing, of course: standing in front of the class talking and writing on the board with chalk or dry-erase markers, or at a lectern watching your PowerPoint slides flicker by. Other options include teaching through individual or group projects, student-centered learning (where students are given responsibility, in essence, for teaching as a vehicle for their own learning), and laboratory exercises. Each of these methods is discussed below, briefly.
Lecturing is a time-honored way of teaching science, and it remains an efficient way of transmitting information. Importantly, it's a style of teaching that most students are familiar with, and therefore comfortable with to varying degrees. Planning lectures can be time consuming during your first years, but once they are planned they can be reused for years, perhaps with only minor updating (depending on how well the lectures worked the first time).
Lecturing with PowerPoint: If you choose to lecture, think carefully about the role of technology in your classroom. Be cautious about using PowerPoint to present information. PowerPoint is too easy; the rate of information transfer can be too fast and your engagement with the material too superficial to engage your students consistently. Chalkboards generally force us to teach at a pace that students can more easily follow.
Carefully designed PowerPoint slides can, however, present important information in a very clear manner, and once you have invested the hours required to make good slides they are relatively easy to update. Your institution may have a cultural or institutional bias for or against the use of PowerPoint in the classroom, so consult with your colleagues before making a commitment to one technology or another.
Other Classroom Technologies: Technologies like smart boards--interactive white boards or computer screens--can enable you to "write" on your presentations on the fly, which helps to give you the best of both worlds--the beauty and clarity of a PowerPoint presentation with the engagement of a chalkboard--but without the logistical barriers a chalkboard or whiteboard can present to the vertically challenged.
If they are available in your classroom, technologies like electronic voting can help engage your students and make them more active. When you ask "Will nitrate spontaneously react with oxygen to form water?" your students must answer electronically before the class moves on. This approach has the additional advantage of allowing you to monitor on the fly how much your students are absorbing.
Research projects and laboratory exercises can capture well the sense of discovery that motivates most scientists. The more open-ended a project is, the more exciting it can be for the student -- and the harder it is to plan, control, and to grade. Projects can be done in lieu of a traditional lab or as an accompaniment to a traditional course. In labs, students can spend a whole semester proposing, implementing, and presenting a project of their choosing or one assigned to them. They might spend the term figuring out the concentration of nitrate in the college pond, for example, tangling along the way with the real-world challenges of accurate assessment. In classes, students can be assigned a topic to investigate--e.g. what is the risk posed by lead to the local community?--and then required to present their answer to the class at the end of the semester.
Some scientists are uncomfortable with the lack of control implicit in this approach. A key to making it work is to be--or become--comfortable with not presenting yourself as an expert all the time. Position yourself, not as a science encyclopedia but as an expert learner. You need to tell your students--and demonstrate--that a large part of your expertise rests in your ability to learn new things efficiently and well. Help them get used to hearing your say "I don't know, what do you think?" There are other, better, ways to win your students' respect than to impress them with how many facts you know. Impress them instead by modeling the inquisitive approach you would like them to take.
Problem-based learning: One well-developed subset of student-centered learning is known as problem-based learning. In this teaching approach, students spend most of the semester solving problems in class. Faculty members do little, or no, lecturing, instead providing students with sets of well-considered problems, and working on them, with students, in groups. As students work together to solve the problems you assign, they gain not just knowledge but new problem-solving skills.
Developing a problem-based course can take lots of time and, for the inexperienced, can be hard to deliver with the right tension (e.g. keeping students engaged and challenged and learning and not slacking off or feeling rushed). A middle-of-the-road alternative that has worked for me is to develop a few topics as problem-based exercises and mix these problem-solving days in with regular lectures. Establish problem-solving groups early in the semester, since they often form the basis of out-of-class study groups.
Make sure groups are balanced with respect to gender and skills. A discussion about implementing problem-based learning in analytical chemistry can be found in this report . This site from Samford University  provides examples, assessment information, and background on problem-based learning.
Take some time to think about how to assess your students, because assessment is the clearest indication of what you value as a teacher. If you claim to value critical thinking you need to give tests and other graded assignments that assess this skill and not rote memorization.
Evaluating student work is time consuming and generally mind numbing, so you want to keep it to a minimum. Look for ways to grade efficiently. Well-written multiple choice questions can test knowledge effectively and are easy to grade, but they take time to write and make cheating easier if you don't take preventive measures. Peer review of lab reports--students evaluating other students, for the benefit of both--can save time. Clear grading rubrics make grading easier and help clarify your expectations.
Some on-line tools can be very effective. Interactive homework problem generators that can generate different--but equivalent--problems for each student and grade student work in real time; students do their homework online, and grading is automatic. Some of the advantages of this approach are obvious; others are less so. In the program I am most familiar with, an adaptable, open-source program called LON-CAPA , students have a set number of tries to get the right answer before they are locked out. This kind of real-time feedback can be a big advantage for your students' learning. These programs also are meticulous about grading things like units and significant figures, details that human graders often overlook. One final advantage: after the deadline, the problem set is no longer accessible; when they miss the deadline it's their problem, not yours.
Whatever decisions you make about assessing your students explain those policies and your rationale in class and on your syllabus. Be explicit. Help students understand what skills you intend to reward.
Start out grading hard, then lighten up. It is always easier to back off than to step up the pressure.
One final note on assessment: grading IS important. It isn't fair to expect your students to take an idealistic attitude towards their work; whether it's med school or grad school, their grades shape their futures.
Some general teaching tips . . .
Be aware of gender and race biases--as well as other biases--in the classroom. Most of us think we don't have any gender and race biases, but there is no shortage of unconscious bias in the classroom, and those people, too, thought they were beyond such things. See this article  for a quick overview of things to guard against.
Protect your time. Students love an "open door" policy, but if you don't set specific office hours your whole day may be spent in impromptu meetings with students. To avoid a steady stream of students "just dropping by," I avoid being at my desk during high traffic times; instead, I'm in the lab doing research. When students enter the lab, I don't slow down. Avoid answering student e-mails instantly; specify a response time--1 to 2 days works well--on your syllabus so students don't come to expect an instant response.
Cheating is very destructive, especially to students who cheat; do not tempt them. Make sure exams and quizzes are proctored and that students spread out in the classroom. Look out over the classroom regularly during an exam or quiz and catch--and hold--students' eyes when they roam. I regularly make multiple copies of the same exam by switching small details in a problem. Open-book exams and out-of-class exams can be useful ways of assessing student knowledge; in my experience, however, they are always accompanied by some degree of cheating or unauthorized collaboration. One solution is to allow some collaboration, but then you have to find some other means to assess how much each student has learned. Establish an anonymous venue where students can report cheating.
Remember that assessment works both ways at most institutions, and plan accordingly. Almost all tenure-and-promotion decisions assess teaching. Student evaluations have been shown to have little merit, but most institutions use them anyway. You can influence these with careful planning; you want your students to like you on evaluation day.
Teaching is an empirical, imminently practical pursuit. Choose a teaching strategy that takes advantage of your strengths and blends with the culture of your students and institution. Avoid the most time-consuming approaches, like service learning, until after tenure. If you're inclined to try to reform that culture, wait until after tenure.
Experiment, and do what works. Every time you teach, try something new, and take the time to write it down and assess (informally but in writing) how well it worked. Assess your own performance, but do not dwell on your own failures; to do so is to assure that they will persist. Record your teaching experiments in your tenure dossier; buttressed by evidence from improved student or faculty evaluations, evidence of creativity in teaching is precisely what tenure and promotion committees are looking for.