I describe two recent projects to test star formation theory using Zeeman observations. First, using Bayesian analysis, the probability distribution function of the magnitude of the total magnetic field strength Bt and its dependence on volume density n(H) were inferred from Zeeman observations of the line-of-sight strengths Bz. The result was that from one molecular cloud to another Bt ranges uniformly between values close to zero and a maximum B0, and that B0 scales as n2/3. Second, observations of the ratio of the mass/flux (M/Φ) between the core and envelope regions of four dark clouds yielded values < 1. All of these results disagree with predictions of the strong magnetic field, ambipolar diffusion driven theory of star formation.