Th e s t ~ t eme nt of the problem which this r esearch isrnc once rned with is to determine whethe r the incidental me thodrnor the systematic method i s mo r e e ffectiv e in teaching sp e llingrnto grade nine Eth i opian students.rnFor this purpose , four sections of grade nine studentsrnwere selected by drawing l ots. By drawing l ots, they wererna ls o ass ~ ed t o ~vo sections in t he c ontrol group a nd thern"rnother two sections in the ex periment a l gr oup . A pre-tes trnwhich was pr epared b y the r esearch er was g iven t o b o th groups.rnThe r esult of the p r e - test indicated that students in bot.rngr oups had similar b a ckgrounds . For thre e and a half mOlths,rnthe students l earned spelling : the c ontrol g roup through thernincident a l me thod i n the normal schedule of t I C school and thernexp erim enta l group through the systematic method on the basisrnof the materia ls f or te a ching spel ling prepared by t h ernreasercher.rnAf t er l e arning spelling f or t hree and a half months, arnpost-test pr ep ared b y the researche r was g iven to the studentsrnof both groups t o de t er min e the hypothes$s ,i. e . ( a) t h e NullrnHypothesis which states tha t there is no sta tist ically significantrndifference between t h e incident a l met ~ od and the sys temarntic method in teaching spelling to gr ade nine Ethiop i a n students,rnand (b) t h e 1.lternate Hypo thesis which s ".; a t es thatrnthere is a statistica lly significant diff er ence b e twe en thernincidental method and the systematic method in teaching spellingrnt o grade nine Ethiopian students.The dat a collected from the pre - test and the po st-tes trnhave be en used for classification a nd e v a l uation to arriverna t a decision .rnOn the basis of a two-ta iled test at a 0 . 01 level ofrnsignifi c ance , we would r eject the Null Hypothesis , if thernvalue of t were greater than t. 995 , wh ich f or (NE + NC - 2 )=rn78 + 80 - 2 = 1 56 degr ee of freedom is 2 .58 . The c omp utedrnvalue of t i s 6 . 08 .rnThus, we c onclude tha t at a 0.01 level of significancernthere is a statistica lly s i gnif i cant differenc e in the perfrnorm ance of the students in the experimental and the controlrngrou ps and we a ccept the ~l ternate Hypothesis at a 0 . 01rnl evel of significa nc e . In addi tion t o this, the analysisrno f the responses of the ques tionnair e shows results similarrnto the t-Tes t which is in f avo ur of the sys t ematic methodrnof teaching spelling .rnOn the basis of these findings, it was conc luded thatrnt h e systema tic me thod is mo r e effe ctive than the incidenta lrnmethod in t eaching spelling to gr ade n ine Ethi opian students.rnRec omme ndations are made on the basis of these findingsrnthat curriculum designers, textbook writers, and educ atorsrnneed to r econs i der their r o l es in promoting the meth od andrnthe proc ess of learning-tea ching spelling and that furtherrnresearch i s ne c essary on the effective t eachin~ of spelling .