I believe in establishing a predictable routine for any class that I teach. At the lower ages, structure is even more critical. Because my students know generally what to expect from class each day, they are less vulnerable to distraction. After a couple of months, the class is almost able to carry out the daily functions without me even saying a word; thus, we can focus on learning physics rather than on what specifically to do.
Now, your routine will almost certainly differ from mine, depending on your personality, your class time, class size, age of student, level, all sorts of things. But I'm asked often enough about what actually happens in class on a day-to-day basis that I think it useful to go through the routine.
In 9th grade, on a typical class day:
We start with a 3 to 4 minute quiz, during which I collect the problem set that was due. (Since I collect the problems from each student's desk personally, it's nearly impossible for a slacker to skate by without me noticing incomplete homework.) The students trade and grade the quizzes.
I take just a few minutes to answer questions and to show the class any information necessary for the day's activity. If we have truly new material, I've already printed and handed out a fact sheet for their reference.
I ask who had the highest quiz score; this student gets to choose the Pandora station that I play during the rest of class.*
* Music is perhaps the most interesting innovation from this past year of teaching freshmen. Since the last half of class almost always involves students working independently as opposed to me talking at the front of the room, there's no reason NOT to put some music on in the background. The quiz is taken that much more seriously, knowing that music selection is the reward for performance. It's amazing to me just how important this reward is to the class. And, woe to the class when someone doesn't complete his problem set. We don't listen to music unless everyone turned in the homework. Peer pressure can be useful...
Students are released to work at their own pace on a set of problems and experiments. Generally, the students solve a problem or a series of problems, as they would on a nightly set. As they finish each problem, they show me their work. If they're wrong, I send them back to their seat to do it right. If they're right, they proceed to the next step.
Usually the "next step" after solving a problem is to head to the back of the room, where the problem is set up as an experiment. The students are asked to perform the experiment to measure whatever they predicted on the in-class problem. For example, the in-class problem might include questions about the motion represented by a velocity-time graph; the experiment would then be to produce the graph with a cart on a track and a motion detector. They show me a printout or a picture of the experimental results... and then they get a new problem to do.
We end class in time to straighten up quickly, and for me to hand out the next day's problem set.
This format has variations... sometimes we do a more traditional laboratory exercise in groups instead of the individually-focused problems and experiments. Sometimes we do test corrections, or problem set review. But the typical quiz - brief talk - independent work model is almost always used.