We test the effect of minimum lot zoning on rural-to-urban land use conversion using Regression Discontinuity Design (RDD), a technique that exploits natural discontinuities in the data to identify causal effects. Observations are drawn from either size of a discontinuous minimum lot size zoning boundary. Using these selected sub-samples, a binary discrete choice model of residential land use change is estimated using parcel-level data and other spatially explicit data from an exurban county that lies on the fringes of Cleveland, Ohio. Results show that controlling for unobserved correlation in the data clearly identifies a negative and significant effect of larger minimum lot size zoning on the probability of conversion to a residential use.