In distributed communication, each transmitter prepares an ensemble of channel codes. To encode a message, a transmitter chooses a channel code individually without sharing the coding choice with other transmitters or with the receiver. Upon receiving the block of channel output symbols, the receiver either decodes the messages of interest if a pre-determined reliability requirement can be met, or reports collision otherwise. Revised from the distributed channel coding theorems in the literature, distributed capacity of a discrete-time memoryless multiple access channel is defined and derived under the assumption that codeword length can be taken to infinity. An improved achievable error performance bound is presented for the case when codeword length is finite.