In modeling online handwritten characters, imaginary strokes have been conveniently generated by connecting ad- jacent real strokes together to form a continuous trajectory. However, this approach causes confusions among charac- ters with similar but actually different trajectories. In this paper, we propose to use Multi-Space probability Distri- bution (MSD) to model imaginary strokes jointly with real strokes. With the proposed MSD, real and imaginary strokes become observations from different probability spaces and they are modeled stochastically. Also, the flexibility in MSD to assign different feature dimensions to each individual space enables us to ignore certain features that can cause singularity problem in modeling. Experimental results ob- tained in handwritten Chinese character recognition indi- cate MSD provides 1.3% - 2.8% character recognition ac- curacy improvement across different recognition systems where MSD significantly improves discrimination among confusable characters with similar trajectories (e.g., ` -b' v.s. and ` -h' ).