Selecting a measuring strain response point as the reference point，a concept named strain power spectrum density transmissibility（SPSDT）is defined. It is theoretically proved that the SPSDT equals to the ratio of strain mode coefficients at the pole of the system. Using this unique property of SPSDT，a series of different reference points are selected to construct the SPSDT matrix，and singular value decomposition is then carried out at the poles of system. It is demonstrated that the first column of the left singular matrix is the strain mode vector so that the operational strain modal parameters of a structure can be identified. The proposed SPSDT based operational strain modal parameter identification procedure does not need an ideal white noise assumption for excitations and only one loading condition is required. The approach is verified by numerical simulation case and a simply supported beam is tested in laboratory. Frequency Domain Decomposition and Stochastic Subspace Identification are compared to prove the effectiveness of the proposed SPSDT based operational strain modal parameter identification. In addition，the influences of data length on the identification results are performed. The results show that the proposed SPSDT based operational strain modal parameter identification can achieve a stable accuracy with only 1 min data and has a good robustness.