In this paper, the complexity and performance of two novel list sphere detector (LSD) algorithms are studied and evaluated in multiple-input multiple-output orthogonal frequency division multiplexing (MIMO-OFDM) system. The LSDs are based on the K-best and the Schnorr-Euchner enumeration (SEE) algorithms. The required list sizes for LSD algorithms are determined for a 2×2 system with 4- quadrature amplitude modulation (QAM), 16-QAM, and 64-QAM. The complexity of the algorithms is compared by studying the number of visited nodes per received symbol vector by the algorithm in computer simulations. The SEE based LSD algorithm is found to be a less complex and a feasible choice for implementation compared to the K-best based LSD algorithm.