This　current　paper　shows　the　asymptotic　normality　for　wavelet　deconvolution　density　estimators,　when　a　density　function　belongs　to　some　L-P(R)　(p　＞　2)　and　the　noises　are　moderately　ill-posed　with　the　index　beta.　The　estimators　include　both　the　linear　and　non-linear　wavelet　ones.　It　turns　out　that　the　situation　for　0　＜　beta　＜=　1　is　more　complicated　than　that　for　beta　＞　1.　(C)　2018　Elsevier　Inc.　All　rights　reserved.