Visible light communications seeks to leverage an unused medium for indoor wireless communications. A major goal is to deliver very high data-rates through LED luminaires to all places where we use lighting. However, the characteristics of LEDs and the nature of indoor lighting conspire to distort the signals. Illumination powers LEDs have low signaling bandwidth and exhibit severe frequency distortion. Their wide dispersion patterns, required for light and signal coverage, also add multipath distortion. Intermittent shadowing results in a wide range of channel characteristics. In this paper we address these challenges with an adaptive receiver. Namely, training is used to identify channel impairments, and our proposed receiver applies specific countermeasures including threshold detection, RAKE reception and adaptive channel equalization. Analysis and simulation demonstrate that our design mitigates distortion problems yielding a performance improvement of 40% to 100% with respect to the current literature in achievable bit-rate depending on the propagation scenario. © 2013 IEEE.