This article establishes an enhanced sliding-mode observer (SMO) with current measurement error compensators for the speed and position estimation in the permanent-magnet synchronous machine (PMSM) sensorless drive scheme. First, hyperbolic tangent switching function-based SMO with a self-adjusted shape coefficient is realized using a fuzzy logic controller (FLC) to not only eliminate the low-pass filter (LPF) and angular compensator in the conventional SMO, but also to reduce chatter caused by the constant boundary layer thickness. On the basis of Lyapunov's second theory, reachable and stable conditions of the enhanced SMO are founded. Then, as the current controller in the PMSM sensorless drive system needs to detect phase currents using current sensors, inevitable measurement errors exist. Two major kinds of current measurement errors, namely, the offset error and the scaling error, are analyzed. Their influences on the current fluctuations and corresponding compensation strategies are given in detail. By integrating current compensators into the enhanced SMO, a practical PMSM sensorless field-oriented control (FOC) scheme is constructed. The effectiveness of the enhanced SMO and current compensators is validated separately by comparative experiments. Final results indicate satisfactory chatter reduction and current ripple suppression performance of the proposed method.