测定了161份不同类型水稻材料的谷粒、糙米的密度及其粒型参数。结果表明，谷粒的密度与谷粒千粒重、谷粒粒长呈极显著正相关，与谷粒长宽比和谷粒百粒体积显著正相关；糙米的密度与糙米千粒重、糙米粒长、糙米百粒体积、糙米粒厚均呈极显著正相关，与糙米宽显著正相关。通径分析表明，谷粒和糙米的千粒重对密度的直接正效应最大；百粒体积和粒宽对密度的直接负效应较大；谷粒其他性状也多是通过谷粒千粒重和百粒体积来间接影响谷粒密度的；而糙米不同性状的间接效应较复杂。谷粒回归模型表明，需增加合理解释谷粒密度变异的预测变量；而糙米回归模型表明，糙米千粒重、粒厚、百粒体积可解释糙米密度因变量的70.2%。

The grain density and grain type characters of grain and brown rice of 161 rice varieties were measured. The results showed that, the grain density was exeremely significant positively correlated with 1 000-grain weight and grain length, significant positively correlated with length-width ratio and 100-grain volume; the density of brown rice was extremely significant positively correlated with 1 000-kernel weight, kernel length, 100-kernel volume and kernel thickness, and significant positively correlated with kernel width. The path analysis showed that, both grain and brown grain’s 1 000-grain weight had the most direct positive effect on grain density, while 100-kernel volume and kernel width had the most direct negative effect on grain density. Most other characters of grain affect grain  density indirectly through 1 000 grain weight and 100-grain volume, while the indirect effect of brown grain density was more complex. The grain regression model indicated that it was necessary to increase the predictive variables to get better prediction result; the brown rice regression model showed that the 1 000-grain weight, grain thickness and 100-grain volume could explain 70.2% of dependent variable.