Seven dilemmas of evaluation and measurement posed by the nature of health education are presented, together with suggestions for their resolution. These include the dilemmas of : 1) rigor of experimental design vs significance or program adaptability; 2) internal validity or "true" effectiveness vs external validity or feasibility; 3) experimental vs placebo effectsl 4) effectiveness vs economy of scale; 5) risk vs payoff; 6) measurement of long-term vs short-term out-comon. Emphasis is placed on the need to develop a more cumulative data base through standardization of measures, replication of experiments in different settings, and better documentation, reporting, and diffusion of experiences in practice.