Built-up edge (BUE) has been noted as a major cause for surface finish deterioration in micromachining processes—even a trace formation of hardened and brittle structure on the tool edge alters the chip load, creates ad hoc and irregular material flow patterns, and results in deposits and smeared regions on the machined surface. To date, few investigations have addressed the formation and effects of BUE in micromachining. This paper is one of the first experimental investigations of the BUE effects on surface quality and its prediction in micromachining. Experiments were conducted on a Haas OM2 CNC milling system with air bearing spindle. The experiments consisted of micromilling 12mm long thin channels on 316L stainless steel plates (30mm×40mm×0.5mm) using uncoated tungsten carbide micromills (from Performance Micro Tools with flat end and Φ0.406mm diameter) at 16 different settings of carefully selected combination of cutting speed and chip load with minimum quantity lubrication (MQL). The surface morphology including BUE distribution, as well as BUE formation on tool surfaces were studied ex situ via optical and scanning electron microscopy. We introduce two metrics to quantify the severity of BUE, namely, BUE density and its distribution entropy. These metrics capture, respectively, the extent and dispersion of BUE on the surface. We also conducted empirical studies to assess the extent to which these quantifiers can determine the variation in surface finish (Sa). Results suggest that the BUE is the major determinant of surface finish besides the chip load effect in micromachining. Statistical Granger-causality tests suggest also that estimation of the BUE density can help with improving the accuracy (R2) of surface finish Sa prediction from a negative value (poor fit) with conventional model to 0.4, i.e., 40% of Sa variation in the micromachining experiments can be attributed to the BUE effect.