We tested five course designs that varied in the structure of daily and weekly active-learning exercises in an attempt to lower the traditionally high failure rate in a gateway course for biology majors. Students were given daily multiple-choice questions and answered with electronic response devices (clickers) or cards. Card responses were ungraded; clicker responses were graded for right/wrong answers or participation. Weekly practice exams were done as an individual or as part of a study group. Compared with previous versions of the same course taught by the same instructor, students in the new course designs performed better: There were significantly lower failure rates, higher total exam points, and higher scores on an identical midterm. Attendance was higher in the clicker versus cards section; attendance and course grade were positively correlated. Students did better on clicker questions if they were graded for right/wrong answers versus participation, although this improvement did not translate into increased scores on exams. In this course, achievement increases when students get regular practice via prescribed (graded) active-learning exercises. (Contains 6 tables and 4 figures.)